lvm allert

[mukesh_a]

we are getting following allert in cluster server ,
vmunix: LVM: Recovered Path (device 0x1f0aa100) to PV 3 in VG 2.
vmunix: LVM: Performed a switch for Lun ID = 0 (pv = 0x0000000048cc5040), from raw device 0x1f0da100 (with priority: 1, and current flags: 0x0) to raw device 0x1f0aa100 (with priority: 0, and current flags: 0x80).
vmunix: LVM: Recovered Path (device 0x1f0aa200) to PV 4 in VG 2.
vmunix: LVM: Performed a switch for Lun ID = 0 (pv = 0x0000000048cbd840), from raw device 0x1f0da200 (with priority: 1, and current flags: 0x0) to raw device 0x1f0aa200 (with priority: 0, and current flags: 0x80).
vmunix: LVM: Recovered Path (device 0x1f0aa300) to PV 5 in VG 2.
vmunix: LVM: Performed a switch for Lun ID = 0 (pv = 0x0000000048cbd080), from raw device 0x1f0da300

couls someone let us know why it,s happening whilw i am not able to get any device related above minnor numbert

[Paul Sperry]

Looks like scsi timeout of some kind. busy disks, bad cables, are possible causes.
What does dmesg tell you?
Also try cstm mstm and xstm

[Calandrello]

Friend
this error occurs how much a LVM way presents imperfection, and the communication passes to be transmitted by the other fiber way. but analyzing this error we can see that he himself ja made recovery. that is ja came back to transmit for the main way.

[Torsten.]

Hi,

looks like your primary path to the storage fails sometimes. Once the path came back, LVM switch back from the alternate path to the primary.
You should investigate the primary path (and all the fiber cables and switches in this path).

[Bill Costigan]

What kind of system and storage?

I've seen similar problems with MSA30s SCSI attached to HPUX.

There is a new scsi driver for HPUX that fixes the problem.

[jitendra_8]

Hello Mukesh
Check the connectivity between system and disk unit.If its ok check the disk I/O load whether too high.If its ok it seems that any of ur disk is about to fail.

[vinod_25]

Hi Mukesh,

The errors were due to LVM switching from its primary link to alternate link may be due to some cabling issue or maintenance.

You can see that the alternate link taking over and says Recovered.

To find the device file... use these simple steps...

1. From the log (device 0x1f0aa100)
2. Eliminate 0x1f so the minor number is 0aa100
3. cd /dev/dsk
4. ll | grep 0aa100
will give you cXtXdX

One Note: you didnot assign points to the above replys; points assignment shows the replier that the requestor has tried his solutions and how much the solution has met his request.

Regards.
Vinod

[Vivekanandan Somanathan]

Hi Mukesh,

Whats your PV TIMEOUT value. If you have default TIMEOUT value, this error may come.

If possible change your TIMEOUT value to 120 and check again.


- Vivek